© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Feature Splitting for GPT2-Small
    5. GPT2-Small
    7. Residual Feature Splitting 12288
    9. 8-RES_FS12288-JB
    11. 452
    Prev
    Next
    INDEX
    Explanations

    details, descriptions, and explanations in technical content

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs12288-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_12288
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    12,288
    Data Type
    torch.float32
    Hook Point
    blocks.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    500
    +0.11
    0.5%
    455
    +0.10
    0.5%
    687
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    500
    +0.11
    0.06
    680
    +0.10
    0.05
    455
    +0.10
    0.06
    Negative Logits
     imperson
    -0.76
    owered
    -0.68
    mouth
    -0.66
     awoke
    -0.66
     bapt
    -0.64
    opped
    -0.64
    anners
    -0.63
     woke
    -0.63
     prank
    -0.60
     baptized
    -0.58
    POSITIVE LOGITS
     below
    1.12
     appendix
    1.06
     detail
    1.03
    below
    0.98
    }}}
    0.97
     subsections
    0.97
    endix
    0.96
     section
    0.96
     details
    0.95
     âĩ
    0.95
    Activations Density 0.413%

    No Known Activations